Metraipolis › archive › dc_28344bc2ab05

Towards Unified Evaluation of Prompt Enhancers for Video Generation — arXiv:2610.11736

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.11736
License (as archived)
CC0-1.0
Archived
2026-10-09T06:30:39.036Z
SHA-256 of the archived text
36722d72045ee525c54c07e722995f0e973c5222210a2488202d94b8223d1fe0
Stable URL of this page
https://metraipolis.com/documento/dc_28344bc2ab05

How to cite

Towards Unified Evaluation of Prompt Enhancers for Video Generation — arXiv:2610.11736 — arXiv (en), https://arxiv.org/abs/2610.11736 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_28344bc2ab05 (Metraipolis archive dc_28344bc2ab05, sha256 36722d72045ee525c54c07e722995f0e973c5222210a2488202d94b8223d1fe0)
BibTeX
@misc{dc28344bc2ab05,
  title = {Towards Unified Evaluation of Prompt Enhancers for Video Generation — arXiv:2610.11736},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.11736},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_28344bc2ab05; sha256 36722d72045ee525c54c07e722995f0e973c5222210a2488202d94b8223d1fe0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11736. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Modern video generators can realize increasingly complex visual narratives, positioning the prompt enhancer (PE) as a critical bridge from concise user instructions and multimodal references to structured cinematic plans. However, existing PE evaluation relies on rendered videos, imposing substantial computational and human costs, slowing PE training and iteration, and conflating PE quality with downstream generator behavior. To address this gap, we introduce PEBench, the first unified benchmark for direct PE evaluation across text-to-video, image-to-video, and reference-to-video prompt enhancement. It comprises 1,100 expert-verified cases and 1,005 visual assets, spanning 35 fine-grained tasks with diverse temporal, cinematic, audiovisual, and multi-reference requirements. In addition, we develop PEBench evaluation, an evidence-grounded framework that combines modality-aware fact extraction with rubric-based assessment across 24 criteria. Our systematic evaluation of representative open- and closed-source PE methods reveals an emerging shift from fine-grained descriptive expansion toward intent-preserving cinematic planning, while the caption-reconstruction and forward-refinement methods show complementary strengths in cinematic coverage and semantic fidelity or internal coherence, respectively. Human validation shows that PEBench scores align closely with expert judgments of enhanced prompts and downstream videos from Wan3.0 and MiniMax-H3, indicating that prompt-level evaluation reliably reflects downstream utility.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.11736 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-09T05:55:07.492Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).